The energy use of data centers is a topic that has received much attention, given that these buildings currently account for 1-2% of global electricity use. Cloud computing holds great potential to reduce data center energy demand moving forward, due to both large reductions in total servers through consolidation and large increases in facility efficiencies compared to traditional local data centers. However, analyzing the net energy implications of shifts to the cloud can be very difficult, because data center services can affect many different components of society’s economic and energy systems. This presentation will summarize a six-month research collaboration with EETD, LBNL’s Computational Research Division, and Northwestern University to begin addressing this net energy analysis challenge.
We present a comprehensive open-access model for assessing the net energy and emissions implications of cloud services in different regions and at different levels of market adoption. The model is first applied to assess a shift in information technology (IT) services from local data centers to the cloud. We estimate the technical potential of cloud-based business software to reduce energy use and greenhouse gas emissions in the United States. Model capability is then expanded to address more complex life cycle studies of broader cloud services that augment or replace traditionally non-IT activities, such as digital versus physical media provision. Our methods and results begin to show how key cloud components affect system-wide energy use and help focus future research towards promoting societal energy and environmental benefits through cloud computing services.
A recording of this seminar is available at: https://vimeo.com/70658689